The mysteries of Galaxies and the Universe is mind boggling. It contains nearly all the stars and planets in the Universe, play host to the supermassive black holes in their centres, and serve as signposts delineating the large-scale cosmic web of dark matter structure. How galaxies were formed is a central question in astronomy. And because galaxies live at the intersection of the study of the structure of the Universe as a whole, understanding galaxy formation also means understanding our own galaxy, the Milky Way, and therefore, our own cosmic history, says Ashoka, in his erudite research, in the weekly column, exclusively for Different Truths.
Galaxies contain nearly all the stars and planets in the Universe, play host to the supermassive black holes in their centres, and serve as signposts delineating the large-scale cosmic web of dark matter structure. How galaxies were formed is a central question in astronomy. And because galaxies live at the intersection of the study of the structure of the Universe as a whole and of the properties of the dark matter, gas, stars, and planets within them, the question is interwoven with many other fields of astronomy. Furthermore, understanding galaxy formation also means understanding our own galaxy, the Milky Way, and, therefore, our own cosmic history.
This is a young field of research. One hundred years ago, astronomers were trying to measure the extent of the Milky Way, and the question whether the Milky Way makes up the entire Universe or other galaxies exist beyond its boundaries was hotly debated. The cosmological framework for understanding galaxy formation and evolution was developed in the 1980s and 1990s, and the first large surveys of the distant Universe were undertaken in the early 2000s. There has been tremendous progress over the past decade, with some of the major questions that astronomers were struggling with now settled. We now have an idea of how galaxies came into being and how they changed over time. It is an incomplete story, with glaring omissions, inconsistencies, and unanswered questions; but a story nonetheless.
Largely thanks to surveys such as the Sloan Digital Sky Survey, we now have a fairly complete census of galaxies in the “nearby Universe”: a loosely defined sphere with a volume of about a billion cubic light-years cantered on our own galaxy. The Sloan Survey has mapped about a million galaxies in this sphere, and we can study their luminosities, colours, morphologies, star formation rates, as well as other properties.
Luminous galaxies show a remarkable regularity in the nearby Universe and can usefully be divided into two basic types. Most stars live in large spiral galaxies, which are characterised by majestic rotating disks of young stars with a dense central bulge of old stars. Our sun resides in such a galaxy: a piece of knowledge that reinforces the notion that “we are not in a special location,” as Copernicus first put forth in 1543. Spiral galaxies continuously form new stars in their arms. The hot, short-lived massive stars that are formed in this process give the disks their characteristic blue colour
The other basic type of galaxy lacks spiral arms, is red, and has stopped forming new stars long ago. Historically, these galaxies are called early-type galaxies since it was once thought that they represent an early stage of galactic evolution; and as often is the case, the name stuck though the interpretation evolved. One of the major results of the Sloan Digital Sky Survey is that these two basic galaxy types are well separated in many projections of the parameters of galaxy mass, kinematics, stellar age, colour, luminosity, and the galactic environment. A galaxy is usually clearly either a spiral galaxy or an early-type galaxy: intermediate types are rare.
Significantly, it has also become clear that all galaxies are embedded in large structures composed of dark matter, some – what confusingly termed “halos.” The mass of a galaxy’s dark matter structure is typically five to ten times greater than the mass of the rest of the galaxy, which means that it is the dark matter that dictates the processes driven by gravity. In many ways, dark matter controls galaxy evolution and the stars are just along for the ride. The nature of dark matter is still elusive be – cause we have not yet identified a dark matter particle (if one even exists). Nevertheless, although we do not yet know what dark matter is, we have a very good idea of where it is. In fact, again using the Sloan Survey, we can now statistically tie galaxies to their dark matter halos and offer a full description of the relation between dark matter and normal matter. This description works remarkably well, and it has been tested by studying the gravitational lensing of faint background galaxies by the dark matter. It turns out that galaxies are also very regular in terms of the relationship between dark and conventional matter: when a galaxy’s mass in normal matter is known, the mass in dark matter can be predicted with an accuracy of about 40 per cent.
Two Basic Types of Galaxies
This regularity in galactic properties makes our task easier: we do not have to decipher the formation history of every individual galaxy. But we do have to understand how the two basic galaxy types came into being and why they are so distinct from one another. And once we understand how typical spiral galaxies were formed, we can apply this lesson to the Milky Way and learn about our own cosmic past.
Our galaxy and the other galaxies in the local volume of the Universe hold important clues to their history. The great majority of stars that have formed in the history of the Universe are still around today, and the present-day appearance of a galaxy is the accumulation of all the things that happened to it over the course of cosmic time. Just like a paleontologist reconstructs a dinosaur from bone fragments, we can use this “fossil evidence” of earlier epochs to reconstruct what galaxies looked like in the past.
From studies of the Milky Way and other nearby galaxies, it appears that the Universe is currently in a much more sedate state than it was in the past. Most galaxies form new stars at a relatively low pace, and so to have built up the vast reservoirs of existing stars that we observe today their formation rates must have been much higher in the past. Studies of the ages of the various components of the Milky Way and its neighbours suggest that the distinction between spiral galaxies and early-type galaxies, which is so characteristic of the distribution of galaxies, may be a transient phenomenon in the history of the Universe. Spiral disks are relatively young and may have arrived on the scene in their present form only in the past five to eight billion years.
Perhaps the most spectacular result of studies of nearby galaxies lies hidden in their outskirts. Sensitive stellar mapping programs of the Milky Way and its neighbour M31 (the Andromeda galaxy) have demonstrated that they are embedded in a vast network of stellar streams, the debris of previous encounters with other galaxies. Such features had first been seen around some galaxies many decades ago, but it is now thought that all galaxies may have vast debris fields around them, although often just below the detection threshold of present-day instrumentation. These streams point toward a violent past when interactions and collisions among galaxies were much more common than they are today.
Galactic paleontology has fundamental limitations, just as looking at dinosaur bones only gives us incomplete and fragmented information on living, breathing dinosaurs. Galactic collisions and mergers erase much of the past history of a galaxy, making it difficult to discern how it was built up. Furthermore, other processes such as bar instabilities and stellar migration gradually change the appearance of galaxies over time even if they do not experience collisions and are instead left to their own devices. As a result, the key building phases of today’s galaxies cannot be deciphered from their present-day appearance alone. Luckily, we are not limited to our own neighbourhood, and we can do something that dinosaur-hunting paleontologists can only dream of.
Owing to the finite speed of light, we can directly observe the past. Looking out into space, we see the moon as it was a second ago, the sun as it was eight minutes ago, and the Andromeda galaxy (the most distant object visible to the unaided eye) as it was 2.5 million years ago. Two and a half million years is only 0.17 per cent of the 14.8-billion-year-old Universe, which is why we (again, somewhat loosely) consider the local volume representative of the present-day Universe.
Using large telescopes on Earth and in space, we are able to detect and study galaxies well beyond the local volume, at distances where the look-back time is a significant fraction of the age of the Universe. Until a few decades ago, we could identify samples of galaxies at distances of about seven billion light-years, allowing us to look back in time about half the age of the Universe. In the mid-1990s, with the combination of Hubble in space and the Keck telescopes on Earth, we learned to take sharp images of galaxies over 95 per cent of the cosmic time, lifting a veil from the early Universe. This frontier is continuously pushed farther into the past, as new detector technology greatly expands the capabilities of existing telescopes. The final space shuttle servicing mission of the Hubble Space Telescope was of particular importance, as the new instruments improved the sensitivity of Hubble by factors of five to twenty.
The Hubble observations, aided by studies at other wavelengths and with large telescopes on Earth, paint a picture of dramatic change. Ten billion years ago galaxies were two to four times smaller than they are today, and yet the rate of star formation was ten times greater. Combining these results, the density of star formation (how many new stars are formed in a fixed region of space within a galaxy) was up to one hundred times greater in these early Universe galaxies than it is in galaxies today. Not surprisingly, the early Universe galaxies also look very different from galaxies today: they are bluer and have a more irregular appearance. The grand spiral galaxies with gently spinning thin disks that are now so ubiquitous were rare in the early Universe.
Star Formation of Early Universe Galaxies
The high star formation rates of early Universe galaxies tell us that they built up rapidly – so rapidly that many doubled their mass in less than a billion years. Eleven billion years ago, the Milky Way was a faint little blob with only 10 per cent of its present-day stellar content but a very large amount of gas: the fuel for star formation. Over the next three to four billion years it proceeded to convert this gas into stars at a ferocious rate, adding about ten times the mass of the sun every year. It then gradually quieted down, became redder as its stellar population age, and settled in its current spiral galaxy morphology about five billion years ago. Over this entire period, Milky Way – like galaxies increased their mass by a factor of ten and grew in size by a factor of two.
Interestingly, some galaxies did not participate in the overall gas feeding frenzy in the young Universe. Despite the availability of large amounts of fuel, about 50 per cent of the most massive galaxies stopped forming new stars as early as ten billion years ago. This strange reluctance to form new stars was already suggested by the old ages of stars in present-day massive galaxies, and it has now been confirmed by direct observations of massive “dead” ancestor galaxies in the young Universe made with the Hubble, Keck, and Gemini telescopes. The structure of these ancestors, as revealed by the Hubble Space Telescope, did yield a surprise. It turns out these galaxies were extremely compact in the past, much smaller than they are today. Remarkably, their sizes increased by a factor of four over the past ten billion years, whereas their masses increased by only a factor of two.
Overall, the epoch around eight to ten billion years ago was characterised by a high degree of diversity. We see very compact “dead” galaxies, large and thick star forming disks, dust-enshrouded collisions, and many other galaxy types. This was also the era when massive black holes at the centres of galaxies were growing rapidly: many galaxies show activity in their nuclei that cannot be explained by star formation and instead reveals the energetic processes associated with black hole growth. This period has been described as “high noon,” the “heyday of galaxy formation,” or – with a nod to our colleagues in the biological sciences – the “cosmic Cambrian,” as the Universe seemed to be experimenting with galaxy shapes and sizes.
Our views of even earlier epochs are necessarily less complete since the feeble light of the galaxies comes from even greater distances. Nevertheless, using the deepest images of the night sky ever obtained astronomers have identified galaxies to within a few hundred million years of the Big Bang and characterised their star formation rates, sizes, and other properties. It now appears that the average star formation rate in the Universe increased rapidly in the first billion years after the Big Bang, after which it had a broad peak and then declined. Interestingly, we can now study galaxies at epochs when the hydrogen that exists in the vast spaces between them was still partially neutral; it is an open question whether the ultraviolet radiation of the hot, massive stars in these early galaxies were responsible for ionizing the Universe.
The Process of Galaxy Formation
As observations of the Universe are providing an increasingly detailed description of the properties of galaxies, the research focus is not only on what happened but also on why it happened. There is broad consensus on the general outline of the process of galaxy formation: gravity and the expansion of the Universe rule the behaviour of dark matter, and give rise to dark matter objects with a distribution of masses that roughly follows a power law (wherein the mass of dark matter can be predicted approximately as a power of the mass of conventional matter). Gas initially follows the distribution of dark matter but then cools and forms stars. The efficiency of this process depends on the dark matter mass, such that the final distribution of the stellar masses of galaxies is not a power law, but has a preferred scale around the mass of the Milky Way.
The details of these processes are fiendishly complex because the dark matter, gas, and stars are all intertwined and continuously changing as structures grow. Furthermore, the relevant physical processes happen on an enormous range of scales, from the distances between galaxies to the central regions of the birth clouds of individual stars: a range that spans thirteen orders of magnitude. To put this challenge into context, it is equivalent to simultaneously understanding the processes that operate on the scale of the Earth-moon system and processes that operate on the scale of the width of a human hair. As if this were not difficult enough, the relevant time scales range from the billions of years of galaxy interactions to the ten-thousand-year free-fall timescale of protostellar clouds.
Despite these seemingly insurmountable challenges, in the past five years, there have been some remarkable successes in modelling the process of galaxy formation. These have mostly come from advanced algorithms running on the world’s fastest computers, a method that has its roots in the first galaxy formation simulations done in the late 1960s. The simulations treat the dark matter and stars as collisionless particles, where a single “particle” usually stands for some ten thousand actual stars. Gas is treated with hydro-dynamical techniques, which simulate the flow of gas and incorporate cooling and heating. The problem of the vast range of scales is ameliorated by adaptively changing the physical scale in the simulation, using a coarse grid for the space in between galaxies and a very fine grid inside the star-forming complexes of spiral galaxies. As even this fine grid can – not capture the formation of individual stars, analytical prescriptions are used for the “subgrid physics”; that is, the processes that happen on scales that are not re – solved by the simulation.
The simulations provide us with movie clips showing the formation of galaxies over the entire history of the Universe, sped up by a factor of 10 to the power 16.They show how little star-forming fragments assemble from gas clouds in the first billion years after the Big Bang, how these fragments grow and merge with one another, how spinning disks form from gas that either condenses gradually out of the dark matter halo or is injected by cold streams, how these disks are destroyed or puffed up in subsequent collisions with other galaxies, and how massive galaxies continue to grow by consuming their little neighbours.
Artificial Galaxies and Simulation
Several results stand out. Perhaps the most impressive is that we now have artificial galaxies living in computers that look a lot like normal spiral galaxies and early-type galaxies in the actual Universe. This is an outstanding achievement: until recently it was not possible to start a simulation shortly after the Big Bang and end up with anything that remotely resembled the Milky Way. In terms of the relevant physical processes, a critical breakthrough was the discovery that gas can get into the central regions of dark matter halos via two paths: it can be shock heated to the virial temperature of the halo followed by gradual cooling, and it can flow directly to the centre along a filament or stream. The simulations have also demonstrated the importance of mergers with small galaxies for the growth of massive early-type galaxies: the cores of massive galaxies form first, and then their outer envelopes are added gradually by accretion. Finally, the simulations consistently demonstrate the overwhelming importance of feedback processes; that is, how much energy is returned to the interstellar medium by newly formed stars and black holes.
These accomplishments come with several crucial asterisks. Many of the key processes, in particular those relating to stellar and black hole feedback, take place on unresolved scales (the subgrid), which means they are essentially free parameters in the simulations. Furthermore, galaxies in our computer simulations simply love forming stars: they are much more efficient at it than the real ones appear to be. We can artificially lower the star formation efficiency by tuning the subgrid parameters, but this is a far cry from understanding the physical processes involved. A possibly related issue is that it has proven difficult to match all observations at once. For example, the simulations that successfully produce early-type galaxies have difficulty producing realistic Milky Way–like galaxies. Finally, the simulations are typically tuned to match the appearance of galaxies in the present-day Universe, and they may not fare as well when compared to the newly available observational data on galaxies at earlier times.
Over the past decade – and even in the last five years – we have made dramatic progress in our understanding of galaxy formation and evolution. We now have a broad idea of the processes that governed the fourteen-billion-year-long home improvement project that was the making of the Milky Way. At the same time, we are only at the beginning: we now have a better idea of the problems we need to solve (such as the too-high star formation efficiency in model galaxies); but that is quite different from actually having the solutions. Additionally, our information on distant galaxies–while greatly improved over the course of the past decade–is still crude by most standards, since we typically only have a few characteristic numbers to work with: galaxies’ luminosities, colours, sizes, and a rough measure of their star formation rates. Perhaps most fundamentally, this review did not touch on the biggest questions of all: until we pin down the nature of dark energy and dark matter, we can hardly claim to understand the formation of structure in the Universe.
As new observing facilities come online and computers and computer algorithms continue to improve in the next decade, and as we gain a better understanding of the physical processes driving galaxy formation through advances in other subfields of astronomy, we can expect further progress. The ALMA facility in Chile will provide us with detailed images of distant galaxies in the light of molecular gas, allowing us to directly connect the existing stars to the fuel for new star formation. The James Web Space Telescope, the successor to the Hubble Space Telescope, will open up the earliest epochs of galaxy formation for study and provide vastly deeper and higher resolution views of distant galaxies than we have access to now. New ground-based telescopes will explore both the large-scale distribution of galaxies and provide high resolution images and spectroscopy of small samples. Finally, the Gaia mission will provide a three-dimensional map of about a billion stars in our own Milky Way, allowing us to piece together our own history via “galactic paleontology” on a vast scale.
©Ashoka Jahnavi Prasad
Photos from the internet.
#Universe #TheMystriesOfTheUniverse #GalacticPaleontology #ArtificialGalaxies #StarFormation #TypeOfGalaxies #MidweekMusing #DifferentTruths
Ashoka Jahnavi Prasad is a physician /psychiatrist holding doctorates in pharmacology, history and philosophy plus a higher doctorate. He is also a qualified barrister and geneticist. He is a regular columnist in several newspapers, has published over 100 books and has been described by the Cambridge News as the ‘most educationally qualified in the world’.